Photochemical generation and 1H NMR detection of alkyl allene oxides in solution

Article abstract of DOI:10.1139/v05-139

Irradiation of substituted 5-alkyl-4,5-epoxyvalerophenones leads to the formation of alkyl allene oxides that, in some cases, are sufficiently long-lived to be detected at room temperature by ¹H NMR spectroscopy. Absolute life-time measurements

Full text of DOI:10.1139/v05-139

1347
1
Photochemical generation and H NMR detection
of alkyl allene oxides in solution1
Leah E. Breen, Norman P. Schepp, and C.-H. Edmund Tan
Abstract: Irradiation of substituted 5-alkyl-4,5-epoxyvalerophenones leads to the formation of alkyl allene oxides that,
1
in some cases, are sufficiently long-lived to be detected at room temperature by H NMR spectroscopy. Absolute life-
time measurements show that the size of the alkyl group has a significant influence on the reactivity of the allene ox-
ide, with tert-butyl allene oxide having a lifetime of 24 h in CD3CN at room temperature that is considerably longer
than the 1.5 h lifetime of the ethyl allene oxide. The allene oxides react rapidly with water to give α-hydroxyketones.
The mechanism involves nucleophilic attack to the epoxide carbon to give an enol, which can also be detected as an
1
intermediate by H NMR spectroscopy.
Key words: allene oxides, mechanisms, absolute reactivity, kinetics, photochemistry.
Résumé : L’irradiation de 5-alkyl-4,5-époxyvalérophénones substituées conduit à la formation d’oxydes d’alkyle et
d’allène qui, dans certains cas, ont des temps de vie suffisamment longs pour qu’on puisse les détecter, à la tempéra-
1
ture ambiante, par spectroscopie RMN du H. Des mesures de temps de vie absolus montrent que la taille du groupe
alkyle joue un rôle important sur la réactivité de l’oxyde d’allène; le temps de vie de l’oxyde de tert-butyle et d’allène,
dans le CD3CN et à la température ambiante, est de 24 h ce qui est beaucoup plus long que celui de l’oxyde d’éthyle
et d’allène qui n’est que d’une heure et demie. Les oxydes d’allène réagissent rapidement avec l’eau pour conduire à la
formation d’α-hydroxycétones. Le mécanisme implique une attaque nucléophile du carbone de d’époxyde pour conduire
1
à un énol qu’on peut aussi détecter comme intermédiaire par spectroscopie RMN du H.
Mots clés : oxydes d’allène, mécanismes, réactivité absolue, cinétique, photochimie.
[Traduit par la Rédaction] Breen et al. 1351
Introduction
ies have shown that the reactions of allene oxides are domi-
nated by rearrangement to cyclopropanones and addition of
nucleophiles to give α substituted ketones.
Allene oxides (1–3) have been studied by theoretical and
experimental chemists for several decades. Much of the orig-
inal interest in these compounds came from their close rela-
tionship with cyclopropanones, which are structural isomers
of allene oxides (4–9). More recently, they have been impli-
cated as intermediates in the conversion of fatty acids to
cyclopentenones and α-ketols in some plants and marine or-
ganisms (10, 11), and have also been used as intermediates
in synthetic organic chemistry (11–20).
While allene oxides are generally considered to be reac-
tive species, some are sufficiently stable to be isolated (21–
24). This has allowed for the spectroscopic analysis of some
allene oxides, and has also led to qualitative conclusions
concerning the effect of structure on allene oxide stability
(1). The chemistry of allene oxides that are too unstable to
be isolated has also been examined through the analysis of
products from reactions in which the allene oxides are pres-
ent as undetected, transient intermediates (1, 2). These stud-
Some absolute rate constants for the reactions of allene
oxides in solution have been measured (25–27), but this re-
mains one area of allene oxide chemistry that has not yet
been extensively examined. Quantitative data concerning the
effect of structure, solvent, and other parameters on the reac-
tivity of allene oxides can be readily obtained from absolute
reactivity measurements, and this information can lead to
mechanistic details that are often difficult to determine from
product-based studies. In the present work, we describe re-
sults concerning the absolute reactivity of some monoalkyl
substituted allene oxides. A key feature of this work is the
use of a photochemical method to rapidly generate these
allene oxides. As a result, more reactive and previously un-
1
detected allene oxides could be examined directly by H
NMR spectroscopy.
Results and discussion
Received 12 October 2004. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on 22 October 2005.
The photochemical method used to generate the allene ox-
ides in this work is based on the Norrish type II reaction of
substituted 4,5-epoxyvalerophenones 1 (eq. [1]). With these
substrates, excitation was expected to lead to efficient triplet
formation followed by rapid intramolecular hydrogen trans-
fer to generate a 1,4-biradical that fragments to give the de-
sired allene oxide 2, together with acetophenone.
L.E. Breen, N.P. Schepp,2 and C.-H. Edmund Tan.
Department of Chemistry, Dalhousie University, Halifax, NS
B3H 4J3, Canada.
1This article is part of a Special Issue dedicated to organic
reaction mechanisms.
2Corresponding author (email: nschepp@dal.ca).
Can. J. Chem. 83: 1347–1351 (2005)
doi: 10.1139/V05-139
© 2005 NRC Canada

1348
Can. J. Chem. Vol. 83, 2005
1
R
O
R
O
Fig. 1. H NMR spectra obtained 4 min after irradiation of pre-
cursors (A) 1a, (B) 1b, and (C) 1c in CD3CN at room tempera-
ture.
H
OH
O
O
OH
hν
Ph
+
R
Ph
O
[1]
Ph
1
2
O
A
H
H
H
1a R = (CH3)3C
2a R = (CH3)3C
2b R = (CH3)2CH
2c R = CH3CH2
2d R = H
1b R = (CH3)2CH
1c R = CH3CH2
1d R = H
Ph
O
H
Initial experiments were carried out using the precursor 1a
(R = (CH3)3C), which would give the tert-butyl allene oxide
2a that has been isolated by other methods and fully charac-
H
B
H
H
1
terized by NMR spectroscopy (23, 28). H NMR analysis
showed that irradiation of 1a (ca. 20% depletion) in CD3CN
at room temperature led to the formation of prominent new
signals3 including doublets at 4.34 ppm (J = 3.2 Hz) and
4.28 ppm (J = 3.2 Hz) and a singlet at 3.53 ppm (Fig. 1A),
as well as a singlet at 2.60 ppm. The signals at 4.34, 4.28,
and 3.53 ppm are the same as those previously assigned to
the tert-butyl allene oxide (23, 28), with the doublets being
assigned to the vinyl protons, and the singlet at 3.50 ppm to
the epoxide proton; the signal at 2.60 ppm corresponds to
the methyl protons of acetophenone. Thus, the NMR spectra
obtained after irradiation clearly showed that fragmentation
leading to allene oxide formation as shown in eq. [1] is oc-
curring.
The same method was applied to the generation of other
allene oxides with less bulky groups at the epoxide carbon
that have until now eluded direct detection. For precursor 1b
(R = (CH3)2CH), the NMR spectrum after irradiation in
CD3CN showed the formation of overlapping doublets at
4.30 ppm and a doublet at 3.57 ppm (Fig. 1B). The signals
at 4.30 and 3.57 ppm are similar to those observed for the
tert-butyl allene oxide, which is consistent with the assign-
ment of these signals to the isopropyl allene oxide 2b. Fur-
thermore, the signal at 3.57 ppm is now a doublet, rather
then the singlet that appears for the tert-butyl allene oxide,
as expected for the isopropyl 2b allene oxide in which the
epoxide proton is located adjacent to the CH of the isopropyl
moiety. For precursor 1c (R = CH3CH2), irradiation led to
the formation of new peaks at 4.29 and 3.71 ppm (Fig. 1C).
Again, these chemical shifts and the nature of the splitting
are consistent with the assignment of these new peaks to the
ethyl allene oxide 2c.
O
H
C
H
H
H
H
4.4
4.2
4
3.8
δ (ppm)
3.6
3.4
Fig. 2. Integrated area of the vinyl protons of the tert-butyl (᭹),
isopropyl (᭿), and ethyl (᭺) allene oxides in CD3CN at 25 °C
measured as a function of time after irradiation of the corre-
sponding precursors. The solid lines represent the line-of-best-fit
calculated using a first-order rate expression.
1.5×105
2×105
0
5×104 1×105
Time (s)
Irradiation of precursor 1d in CD3CN led to no detectable
amounts of the parent allene oxide 2d. A signal at 2.60 ppm
owing to the formation of acetophenone was observed,
which suggests that allene oxide formation as outlined in
eq. [1] had occurred. Thus, our inability to detect allene ox-
ide 2d is likely due to its high reactivity. To further establish
that 2d was indeed generated, irradiation of 1d was carried
out with 5% D2O in acetonitrile. Under these conditions, a
prominent singlet at 4.01 ppm and triplet at 2.09 ppm were
observed. These signals correspond to 1-d-3-deuteroxy-2-
propanone (29), which is the expected product upon addition
of D2O to allene oxide 2d. Thus, the parent allene oxide is
generated upon irradiation of 1d, but even in dry acetonitrile
it is too short-lived to be detected under our reaction condi-
tions. Since our experimental conditions involved 5 min of
1
irradiation followed by ca. 3 min to acquire the initial H
NMR spectrum, the lifetime of the parent allene oxide must
be less 8 min in neat CD3CN at room temperature.
Not surprisingly, the allene oxides 2a–2c in neat CD3CN
at room temperature also were not stable, as illustrated by
the decrease in the integrated areas of the H NMR signals
1
of the vinyl protons as a function of time in Fig. 2. These
data can in principle be used to determine the absolute reac-
tivity of the allene oxides; however, given the difficulties in
measuring accurate values for the integrated areas, the data
were not sufficiently good to distinguish between a first-
order decay or a second-order decay of the allene oxides.
Nonetheless, the data clearly show a substantial difference in
the reactivity of the allene oxides as a function of the size of
1
3 Other weak resonances owing to unidentified by-products were also evident in the H NMR spectrum.
© 2005 NRC Canada

Breen et al.
1349
Table 1. First-order rate constants for the
decay of allene oxides 2a–2d in CD3CN (25 °C).
Fig. 3. Rate constants for the decay of the tert-butyl (᭹),
isopropyl (᭺), and ethyl (᭿) allene oxides in CD3CN as a func-
tion of D2O content (25 °C).
R
kdecay (s–1)
0.006
0.005
0.004
0.003
0.002
0.001
0
tert-Butyl (2a)
Isopropyl (2b)
Ethyl (2c)
(1.2 0.2) × 10–5
(2.4 0.4) × 10–5
(1.9 0.2) × 10–4
>10–3
H (2d)
the alkyl group, with the isopropyl allene oxide 2b being
considerably more reactive than the tert-butyl allene oxide
2a, and the ethyl allene oxide 2c showing even greater reac-
tivity. This order of reactivity as a function of the size of the
alkyl group is nicely consistent with conclusions made pre-
viously (2, 24) that bulky groups directly attached to the
epoxide ring serve to enhance the kinetic stability of allene
oxides.
0
10
20
30
40
50
% D2O in CD3CN
While the decay traces in Fig. 2 fit equally well to first-
and second-order rate expressions, a measure of the differ-
ences in reactivity can be estimated by comparing rate con-
stants obtained using a first-order fit (Table 1). These results
show that the tert-butyl allene oxide 2a and the isopropyl
allene oxide decay slowly with rate constants of 1.2 × 10–5
and 2.4 × 10–5 s–1, respectively, which differ only by a factor
of two. On the other hand, the ethyl allene oxide 2c decays
about an order of magnitude more quickly with a rate con-
stant of 1.9 × 10–4 s–1. In addition, based on our inability to
detect the parent allene oxide, we can estimate that it decays
with a rate constant >10–3 s–1. The parent allene oxide there-
fore decays at least 10 times more rapidly than the ethyl de-
rivative, and at least 100 times more rapidly than the bulky
tert-butyl and isopropyl allene oxides.
order rate constants that are plotted as a function of D2O
content in Fig. 3. These data clearly show that all three
allene oxides decayed considerably more quickly upon the
addition of water; however, the reactivity of the allene ox-
ides toward D2O was strongly dependent on the nature of
the alkyl substituent. Thus, in the presence of the tert-butyl
group, the rate constants for allene oxide decay hardly
changed at low D2O contents, and the tert-butyl allene oxide
could still be easily observed in 40% D2O where it decayed
with a rate constant of 6.7 × 10–4 s–1. This is in contrast to
the results obtained with the isopropyl allene oxide whose
decay increased dramatically even in the presence of small
amounts of D2O, and could no longer be observed at D2O
contents >10%. The ethyl allene oxide was even more reac-
tive, with the rate constant for decay increasing 10-fold upon
going from neat CD3CN to 3% D2O in CD3CN. At higher
D2O contents, the ethyl allene oxide could no longer be ob-
served.
The actual reaction of the allene oxides in neat CD3CN re-
mains unclear. One possibility is that the allene oxides rear-
1
range to the substituted cyclopropanones. However, no H
NMR signals corresponding to cyclopropanones could be
observed after complete decay of the allene oxides, nor
could signals owing to the alkene that would be generated
by rapid loss of CO from the cyclopropanone. In fact, no
distinct signals were observed to grow as the allene oxides
decayed, suggesting that the allene oxides may polymerize
or react with solvent to give material difficult to character-
ize.
Nucleophilic addition of D2O to the allene oxides was ex-
pected to lead to the formation of the corresponding α-
1
hydroxyketone, and this material was indeed observed by H
NMR as the final stable product in the reaction of the allene
oxides with D2O. Two mechanisms for this nucleophilic ad-
dition can be envisioned: addition can occur via an SN1-like
mechanism characterized by initial breaking of the epoxide
C—O bond followed by addition of the nucleophile, or via
an SN2-like mechanism where the nucleophile directly adds
to the intact epoxide ring (eq. [2]). In both cases, the initial
product is an enol that then tautomerizes to the more stable
ketone form. The formation of the enol intermediate is sup-
Allene oxides are well-known to react with nucleophiles
to give α substituted ketones (2), but little is known about
the absolute rate constants associated with this nucleophilic
addition process. In the present work, the effect of adding
D2O was examined by monitoring the decay of the allene
oxides signals in various D2O–CD3CN mixtures. Since D2O
was expected to add rapidly to the allene oxides, irradiation
1
ported by the sequence of H NMR spectra in Fig. 4. Fig-
1
ure 4A shows the isopropyl allene oxide after irradiation in
CD3CN, but before the addition of D2O. Five minutes after
the addition of 10% D2O, the allene oxide signals were re-
duced (Figs. 4B and 4C) and eventually replaced by reso-
nances at 3.9 and 4.1 ppm that correspond to the expected
(30) chemical shifts of the vinyl protons of the enol interme-
diate. A doublet at 3.65 ppm was also observed that can be
assigned to the proton on the carbon bearing the hydroxy
group of the enol. The enol signals were then replaced by a
of the precursors was carried out in neat CD3CN, and an H
NMR spectrum obtained to ensure that allene oxide forma-
tion had occurred. An appropriate amount of D2O was then
added, followed rapidly by the acquisition of a series of
spectra. In this way, allene oxide decaying with rate con-
stants less than 0.01 s–1 could readily be monitored.
In these experiments, the reduction in the intensity of the
vinyl protons of the allene oxides as a function of time fit
well to a first-order rate expression giving the pseudo-first-
© 2005 NRC Canada

1350
Can. J. Chem. Vol. 83, 2005
1
Fig. 4. H NMR spectra obtained after irradiation of 1b in
ity upon going from tert-butyl to ethyl should be observed.
As shown in Fig. 3, a fairly dramatic structural effect is ob-
served, which leads to the conclusion that nucleophilic addi-
tion occurs via the SN2 mechanism.
CD3CN, and 5, 19, and 100 min after the addition of 10% D2O.
O
H
H
D
D (100 min)
C (19 min)
OD
H
Conclusions
H
Our results clearly show that photogeneration of allene
oxides by the Norrish type II fragmentation of 4-alkyl-4,5-
epoxyvalerophenones leads to formation of simple alkyl sub-
stituted allene oxides. Since the allene oxides are generated
rapidly, this method allows for the detection of allene oxides
that had not previously been observed, and for absolute rate
constants to be measured. The use of this photoreaction
should be compatible with the generation of broader series
of simple alkyl allene oxides, and also with the use of a fast
reaction technique like flash photolysis to examine the reac-
tivity of very short-lived allene oxides. Experiments to ad-
dress those issues are currently under way.
OD
H
H
OD
H
H
B (5 min)
A (0 min)
O
H
H
H
H
4.4
4.2
4
3.8
3.6
Experimental
δ (ppm)
Preparation of precursors
Precursors 1a–1d were prepared by addition of the enolate
of acetophenone to the appropriate 1-bromo-2-alkene deriva-
tive to give the corresponding 1-phenyl-pent-4-en-1-one, fol-
lowed by epoxidation using m-chloroperbenzoic acid. In
general, the procedure involved the addition of 1 equiv. of
lithium diisopropylamine to acetophenone (1.5 g) in dry
THF (15 mL) under an argon atmosphere at –78 °C and stir-
ring for 20 min. One equivalent of 1-bromo-4,4-dimethyl-2-
pentene (32) (for 1a), 1-bromo-4-methyl-2-pentene (33) (for
1b), 1-bromo-2-pentene (33) (for 1c), or 1-bromo-2-propene
(Aldrich) (for 1d) was then added, and the mixture allowed
to warm up to room temperature over a period of 20 h. The
solution was diluted with ether and washed with saturated
aqueous ammonium chloride and sodium chloride. The or-
ganic layer was dried and the solvent removed under re-
duced pressure. The crude enone was purified by silica gel
column chromatography using 50% dichloromethane in hex-
anes as the eluant.
O
R
SN2
SN1
D
O
O
OD
[2]
OD
D
H2C
D
OD
R
R
O-
R
+
D
O
D
doublet at 3.9 ppm (Fig. 4D), attributed to the proton at the
α position of the expected α-hydroxyketone product4 (29).
While formation of the enol intermediate is consistent
with the mechanism shown in eq. [2], its observation was
surprising given the high reactivity of most simple enols
(31). Two factors may contribute to the unusually long life-
time. First, most absolute rate constant measures for the
ketonization of enols have been measured in wholly aqueous
solutions; in the 10% aqueous conditions used to observe the
enol in Fig. 2B, the enol ketonization process may be some-
what slowed. In fact, in experiments carried out at higher
water contents of 30%, the enol could no longer be ob-
served, despite the more rapid disappearance of the allene
oxide. Second, the α-hydroxy group may play a role in re-
ducing the reactivity of the enol form by hydrogen bonding.
Such hydrogen bonding may reduce the electron density at
the enolic OH, thus reducing the overall electrophilicity and
the rate constant for addition of a proton to the enolic car-
bon–carbon double bond.
Epoxidation was carried out by adding 1.5 equiv. of puri-
fied m-chloroperbenzoic acid to the enone in CH2Cl2. The
mixture was stirred for 1.5 h at room temperature. After
washing with 5% sodium hydroxide and evaporation of the
solvent, the crude product was purified by silica gel column
chromatography (20% ethyl acetate in hexanes).
1-Phenyl-6,6-dimethyl-4,5-epoxyheptan-1-one (1a)
1H NMR (ppm) δ: 7.45–8.05 (multiplet, 5H), 3.14 (t, 2H),
2.92–2.97 (m, 1H), 2.56 (d, 1H), 2.08–2.22 (m, 1H), 1.77–
1.91 (m, 1H), 0.92 (s, 9H). 13C NMR (ppm) δ: 25.7, 26.5,
30.6, 54.5, 67.1, 127.9, 128.5, 133.0, 136.6, 199.0. HR-MS
calcd.: 232.1463; found: 232.1459.
The effect of structure on the absolute reactivity of the
allene oxides with water can also provide some insight into
the mechanism of the nucleophilic addition reaction. In par-
ticular, rate constants for addition of the nucleophile in the
SN2 mechanism should show a dependence on structure,
while in the SN1-like mechanism, little difference in reactiv-
1-Phenyl-6-methyl-4,5-epoxyheptan-1-one (1b)
1H NMR (ppm) δ: 7.45–8.05 (multiplet, 5H), 3.15 (t, 2H),
2.85–2.91 (m, 1H), 2.53–2.57 (d of d, 1H), 2.10–2.23 (m,
1H), 1.78–1.92 (m, 1H), 1.45–1.62 (m, 1H), 1.01 (d, 3H),
4 The final product also shows the CH2D protons as a triplet at 2.1 ppm and the methyl groups as doublets at 1.0 and 0.7 ppm.
© 2005 NRC Canada

Breen et al.
1351
0.96 (d, 3H). 13C NMR (ppm) δ: 18.3, 18.9, 26.4, 30.3, 34.5,
56.8, 64.5, 127.9, 128.5, 133.0, 136.6, 199.0. HR-MS calcd.:
218.1307; found: 218.1299.
References
1. T.H. Chan and B.S. Ong. Tetrahedron, 36, 2269 (1980).
2. T.H. Chan and B.S. Ong. J. Org. Chem. 43, 2994 (1978).
3. P.J. Stang. In The chemistry of ethers, crown ethers, hydroxyl
groups and their sulphur analogues. Vol. 2. Edited by S. Patai.
Wiley, Chichester, UK. 1980. p. 859.
1-Phenyl-4,5-epoxyheptan-1-one (1c)
1H NMR (ppm) δ: 7.45–8.05 (multiplet, 5H), 3.15 (t, 2H),
2.82–2.87 (m, 1H), 2.71–2.77 (m, 1H), 2.12–2.25 (m, 1H),
1.78–1.91 (m, 1H), 1.53–1.64 (m, 2H), 1.00 (t, 3H). 13C
NMR (ppm) δ: 9.7, 24.9, 26.3, 34.5, 57.5, 60.2, 127.9,
128.5, 133.0, 136.6, 199.1. HR-MS calcd.: 204.1150; found:
204.1146.
4. A. Liberles, A. Greenberg, and A. Lesk. J. Am. Chem. Soc.
94, 8685 (1972).
5. M.E. Zandler, C.E. Choc, and C.K. Johnson. J. Am. Chem.
Soc. 96, 3317 (1974).
6. J.V. Ortiz. J. Org. Chem. 48, 4744 (1983).
7. B.A. Hess and J.K. Cha. THEOCHEM, 100, 11 (1993).
8. B.A. Hess, U. Eckart, and J. Fabian. J. Am. Chem. Soc. 120,
12310 (1998).
1-Phenyl-4,5-epoxypentan-1-one (1d) (34)
1H NMR (ppm) δ: 7.45–8.05 (multiplet, 5H), 3.20 (t, 2H),
3.10–3.17 (m, 1H), 2.81 (t, 1H), 2.55–2.58 (m, 1H), 2.16–
2.24 (m, 1H), 1.80–1.89 (m, 1H). 13C NMR (ppm) δ: 26.8,
34.4, 47.3, 51.5, 127.9, 128.5, 133.1, 136.6.
9. B.A. Hess, Jr., L. Smentek, A.R. Brash, and J.K. Cha. J. Am.
Chem. Soc. 121, 5603 (1999).
10. N. Tijet and A.R. Brash. Prostaglandins Other Lipid Media-
tors, 68–69, 423 (2002).
11. E.J. Corey, K. Ritter, M. Yus, and C. Najera. Tetrahedron Lett.
28, 3547 (1987).
12. J. Grimaldi, M. Malacria, and M. Bertrand. Bull. Soc. Chim.
Fr. 7–8, 1731 (1975).
Irradiation of precursors and determination of rate
constants
About 10 mg of 1a–1d were dissolved in 0.750 mL of
CD3CN (Aldrich) in a quartz cuvette (3 mm × 1 cm × 3 cm).
The solution was bubbled with nitrogen for 10 min and then
irradiated with 308 nm light (60 mJ/pulse, 8 ns pulse) from a
Lambda-Physik excimer laser operating at 5 Hz for 5–
10 min or with a high-pressure mercury lamp for 15–30 min.
Immediately after irradiation, the solution was transferred to
an NMR tube and the spectrum obtained using a Bruker AC-
250 MHz spectrometer (probe temperature 25 °C). To obtain
kinetic data, NMR spectra were obtained at specific time in-
tervals after irradiation. The time intervals varied depending
on the lifetime of the allene oxides, but generally six to ten
spectra were obtained. The peak area or intensity corre-
sponding to the vinyl protons of the allene oxides was mea-
sured in each spectrum (relative to a small amount of
CH2Cl2 added as an internal standard). These areas or inten-
sities were plotted as a function of time, and rate constants
extracted by nonlinear least-squares fitting of the data to a
first-order exponential equation.
13. M. Bertrand, J.P. Dulcere, G. Gil, J. Grimaldi, and P.
Sylvestre-Panthet. Tetrahedron Lett. 3305 (1976).
14. M. Bertrand, J.P. Dulcere, and G. Gil. Tetrahedron Lett. 21,
4271 (1980).
15. S.J. Kim and J.K. Cha. Tetrahedron Lett. 29, 5613 (1988).
16. I. Erden, F.P. Xu, J. Drummond, and R. Alstad. J. Org. Chem.
58, 3611 (1993).
17. I. Erden, F.P. Xu, and W.G. Cao. Angew. Chem. Int. Ed. Engl.
36, 1516 (1997).
18. M.M. Kabat. J. Org. Chem. 60, 1823 (1995).
19. M. Shipman, H.R. Thorpe, and I.R. Clemens. Tetrahedron, 55,
10845 (1999).
20. H. Xiong, R.P. Hsung, C.R. Berry, and C. Rameshkumar. J.
Am. Chem. Soc. 123, 7174 (2001).
21. J.K. Crandall, W.W. Conover, J.B. Komin, and W.H.
Machleder. J. Org. Chem. 39, 1723 (1974).
22. I. Erden and J. Drummond. Tetrahedron Lett. 34, 1255 (1993).
23. T.H. Chan and B.S. Ong. Tetrahedron Lett. 17, 3253 (1976).
24. R.O. Camp and F.D. Greene. J. Am. Chem. Soc. 90, 7349
(1968).
25. A.R. Brash. J. Am. Chem. Soc. 111, 1891 (1989).
26. A.R. Brash, S.W. Baertschi, C.D. Ingram, and T.M. Harris.
Proc. Natl. Acad. Sci. U.S.A. 85, 3382 (1988).
27. M.D. Clay, J. Durber, and N.P. Schepp. Org. Lett. 3, 3883
(2001).
For studies involving the addition of water, irradiation of
the precursor was carried out in the same way in neat
CD3CN. The 1H NMR spectrum was then obtained to ensure
allene oxide formation had taken place. A desired amount of
D2O was then added, followed by the acquisition of further
NMR spectra at appropriate time intervals. Rate constants
were determined in the same way as described in the previ-
28. T. Konoike, T. Hayashi, and Y. Araki. Tetrahedron: Asymme-
try, 5, 1559 (1994).
29. J.A. Katzenellenbogen and S.B. Bowlus. J. Org. Chem. 38,
627 (1973).
30. B. Capon, D.S. Rycroft, T.W. Watson, and C. Zucco. J. Am.
Chem. Soc. 103, 1761 (1981).
1
ous section for reaction in neat CD3CN. In all cases, the H
NMR spectra of the final stable product matched closely the
published spectra of the corresponding α-hydroxy ketones
(29).
31. Y. Chiang, A.J. Kresge, and N.P. Schepp. J. Am. Chem. Soc.
111, 3977 (1989).
32. R.W. Hoffmann, H. Brinkman, and G. Frenking. Chem. Ber.
123, 2387 (1990).
Acknowledgements
33. D. Seebach and L. Widler. Helv. Chim. Acta, 65, 1972 (1982).
34. P. Crotti, V. Di Bussolo, L. Favero, F. Macchia, M. Pineschl,
and E. Napolitano. Tetrahedron, 55, 5853 (1999).
The authors thank the Natural Sciences and Engineering
Research Council of Canada (NSERC) and PetroCanada
(Young Innovators Award) for financial support.
© 2005 NRC Canada